Home > Portfolio > PINTEREST-STYLE TEXT-TO-IMAGE APP
AI OpenAI MongoDB React

PINTEREST-STYLE TEXT-TO-IMAGE APP

React-Node mini-project generating pictures via OpenAI API with MongoDB storage.

PINTEREST-STYLE TEXT-TO-IMAGE APP Hero Image

A mini-project built to generate creative images from text prompts using OpenAI's DALL-E, organized in an interactive Pinterest-style layout.

"We built a responsive image grid that saves generated images to MongoDB, allowing users to save, catalog, and share creations."

The Approach

Users enter text descriptions inside the React client. The Node.js backend sanitizes the prompt, forwards it to the OpenAI API, receives the image URL, and commits the metadata to a MongoDB cluster. The client renders images inside a responsive masonry layout resembling Pinterest.

Impact

Demonstrates fast API integration and clean masonry rendering, offering a platform to generate and share generative artwork.

System Architecture

[MASONRY GRID LAYOUT]
React.js App with CSS Grid Masonry rendering
▼
[IMAGE INFERENCE CONTROLLER]
Express Backend / OpenAI DALL-E 3 API / Cloudinary Storage Uploader
▼
[META DATA CLUSTER]
MongoDB Cloud Cluster (Prompts & CDN Image URL References)